On Tue, 24.09.13 11:09, Evan Callicoat ([email protected]) wrote: > > On 09/24/2013 08:27 AM, Colin Walters wrote: > > On Mon, 2013-09-23 at 21:01 -0500, Evan Callicoat wrote: > > > >> Ideally in the long-term, applications which rely on SHELL being set > >> should be fixed to just grab it from getpwnam() or similar, > >> but until that becomes more common, I propose this simple change to > >> make user sessions a little bit nicer out of the box. > > > > Right; this is exactly the kind of thing that would be fixed by having > > logind run the PAM stack and pass the environment/state to systemd-user. > > > > Agreed. I just jumped in here and haven't caught up on all the recent > conversations, so apologies if this is a well-known discussion already. > Is the current plan to have logind run the PAM stack, or is it just an > idea waiting for traction? If it's imminent, no need to merge in my > proposed patch. Otherwise, I still think it'd be a good interim tweak.
logind just starts/stops [email protected] for each user as long as there is a session of that user logged in. [email protected] uses PAMName= as a setting which is how the PAM logic is enabled for it. However, $SHELL is not set by PAM, it's set by something else (login? the shell? dunno...) Lennart -- Lennart Poettering - Red Hat, Inc. _______________________________________________ systemd-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/systemd-devel
